Pest Incidence Prediction in Organic Banana Crops with Machine Learning Techniques

• Autores: Almeyda E., Paiva J., Ipanaque W.
• Formato: artículo
• Fecha de Publicación: 2020
• Institución: Consejo Nacional de Ciencia Tecnología e Innovación
• Repositorio: CONCYTEC-Institucional
• Lenguaje: inglés
• OAI Identifier: oai:repositorio.concytec.gob.pe:20.500.12390/2472
• Enlace del recurso: https://hdl.handle.net/20.500.12390/2472; https://doi.org/10.1109/EIRCON51178.2020.9254034
• Nivel de acceso: acceso abierto
• Materia: trips; binary classification; logistic regression; machine learning; organic banana; pest; support vector machine; http://purl.org/pe-repo/ocde/ford#4.01.01

One of the main problems organic banana crops is the presence of pests, affecting crop yield, post-harvest and export fruit quality. In Piura (Peru), pests with the greatest presence are Thrips, Squamas, Black Weevil, etc. This article describes the development of a prediction model, based on a supervised machine learning algorithm: Logistic Regression and Support Vector Machine, which will estimate the future level of incidence (low and medium) of a specific pest. The model was designed including the input data (climate) that were obtained from a network of IoT sensors in-situ in the banana crop, and output data (level of incidence) that was collected with manual record and visual inspection. The model developed can predict pest incidence at 79% accuracy (with test data). These first results show feasibility to estimate in advance the incidence of pests in that crop. Future implementation of the model would help to farmers improving the pest management to their crops, increasing the production and quality of the product. © 2020 IEEE.